Dataset Information

Uncertainty-aware deep-learning model for prediction of supratentorial hematoma expansion from admission non-contrast head computed tomography scan.

ABSTRACT: Hematoma expansion (HE) is a modifiable risk factor and a potential treatment target in patients with intracerebral hemorrhage (ICH). We aimed to train and validate deep-learning models for high-confidence prediction of supratentorial ICH expansion, based on admission non-contrast head Computed Tomography (CT). Applying Monte Carlo dropout and entropy of deep-learning model predictions, we estimated the model uncertainty and identified patients at high risk of HE with high confidence. Using the receiver operating characteristics area under the curve (AUC), we compared the deep-learning model prediction performance with multivariable models based on visual markers of HE determined by expert reviewers. We randomly split a multicentric dataset of patients (4-to-1) into training/cross-validation (n = 634) versus test (n = 159) cohorts. We trained and tested separate models for prediction of ≥6 mL and ≥3 mL ICH expansion. The deep-learning models achieved an AUC = 0.81 for high-confidence prediction of HE_≥6 mL and AUC = 0.80 for prediction of HE_≥3 mL, which were higher than visual maker models AUC = 0.69 for HE_≥6 mL (p = 0.036) and AUC = 0.68 for HE_≥3 mL (p = 0.043). Our results show that fully automated deep-learning models can identify patients at risk of supratentorial ICH expansion based on admission non-contrast head CT, with high confidence, and more accurately than benchmark visual markers.

SUBMITTER: Tran AT

PROVIDER: S-EPMC10847454 | biostudies-literature | 2024 Feb

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Uncertainty-aware deep-learning model for prediction of supratentorial hematoma expansion from admission non-contrast head computed tomography scan.

Tran Anh T AT Zeevi Tal T Haider Stefan P SP Abou Karam Gaby G Berson Elisa R ER Tharmaseelan Hishan H Qureshi Adnan I AI Sanelli Pina C PC Werring David J DJ Malhotra Ajay A Petersen Nils H NH de Havenon Adam A Falcone Guido J GJ Sheth Kevin N KN Payabvash Seyedmehdi S

NPJ digital medicine 20240206 1

Hematoma expansion (HE) is a modifiable risk factor and a potential treatment target in patients with intracerebral hemorrhage (ICH). We aimed to train and validate deep-learning models for high-confidence prediction of supratentorial ICH expansion, based on admission non-contrast head Computed Tomography (CT). Applying Monte Carlo dropout and entropy of deep-learning model predictions, we estimated the model uncertainty and identified patients at high risk of HE with high confidence. Using the ...[more]

PMID: 38321131

Similar Datasets

Project description:Background and purposeHematoma expansion (HE) is a critical event following acute intracerebral hemorrhage (ICH). We aimed to construct a non-contrast computed tomography (NCCT) model combining clinical characteristics, radiological signs, and radiomics features to predict HE in patients with spontaneous ICH and to develop a nomogram to assess the risk of early HE.Materials and methodsWe retrospectively reviewed 388 patients with ICH who underwent initial NCCT within 6 h after onset and follow-up CT within 24 h after initial NCCT, between January 2015 and December 2021. Using the LASSO algorithm or stepwise logistic regression analysis, five models (clinical model, radiological model, clinical-radiological model, radiomics model, and combined model) were developed to predict HE in the training cohort (n = 235) and independently verified in the test cohort (n = 153). The Akaike information criterion (AIC) and the likelihood ratio test (LRT) were used for comparing the goodness of fit of the five models, and the AUC was used to evaluate their ability in discriminating HE. A nomogram was developed based on the model with the best performance.ResultsThe combined model (AIC = 202.599, χ2 = 80.6) was the best fitting model with the lowest AIC and the highest LRT chi-square value compared to the clinical model (AIC = 232.263, χ2 = 46.940), radiological model (AIC = 227.932, χ2 = 51.270), clinical-radiological model (AIC = 212.711, χ2 = 55.490) or radiomics model (AIC = 217.647, χ2 = 57.550). In both cohorts, the nomogram derived from the combined model showed satisfactory discrimination and calibration for predicting HE (AUC = 0.900, sensitivity = 83.87%; AUC = 0.850, sensitivity = 80.10%, respectively).ConclusionThe NCCT-based model combining clinical characteristics, radiological signs, and radiomics features could efficiently discriminate early HE, and the nomogram derived from the combined model, as a non-invasive tool, exhibited satisfactory performance in stratifying HE risks.

Project description:ImportanceHematoma expansion is a potentially modifiable predictor of poor outcome following an acute intracerebral hemorrhage (ICH). The ability to identify patients with ICH who are likeliest to experience hematoma expansion and therefore likeliest to benefit from expansion-targeted treatments remains an unmet need. Hypodensities within an ICH detected by noncontrast computed tomography (NCCT) have been suggested as a predictor of hematoma expansion.ObjectiveTo determine whether hypodense regions, irrespective of their specific patterns, are associated with hematoma expansion in patients with ICH.Design, setting, and participantsWe analyzed a large cohort of 784 patients with ICH (the development cohort; 55.6% female), examined NCCT findings for any hypodensity, and replicated our findings on a different cohort of patients (the replication cohort; 52.7% female). Baseline and follow-up NCCT data from consecutive patients with ICH presenting to a tertiary care hospital between 1994 and 2015 were retrospectively analyzed. Data analyses were performed between December 2015 and January 2016.Main outcomes and measuresHypodensities were analyzed by 2 independent blinded raters. The association between hypodensities and hematoma expansion (>6 cm3 or 33% of baseline volume) was determined by multivariable logistic regression after controlling for other variables associated with hematoma expansion in univariate analyses with P ≤ .10.ResultsA total of 1029 patients were included in the analysis. In the development and replication cohorts, 222 of 784 patients (28.3%) and 99 of 245 patients (40.4%; 321 of 1029 patients [31.2%]), respectively, had NCCT scans that demonstrated hypodensities at baseline (κ = 0.87 for interrater reliability). In univariate analyses, hypodensities were associated with hematoma expansion (86 of 163 patients with hematoma expansion had hypodensities [52.8%], whereas 136 of 621 patients without hematoma expansion had hypodensities [21.9%]; P < .001). The association between hypodensities and hematoma expansion remained significant (odds ratio, 3.42 [95% CI, 2.21-5.31]; P < .001) in a multivariable model; other independent predictors of hematoma expansion were a CT angiography spot sign, a shorter time to CT, warfarin use, and older age. The independent predictive value of hypodensities was again demonstrated in the replication cohort (odds ratio, 4.37 [95% CI, 2.05-9.62]; P < .001).Conclusion and relevanceHypodensities within an acute ICH detected on an NCCT scan may predict hematoma expansion, independent of other clinical and imaging predictors. This novel marker may help clarify the mechanism of hematoma expansion and serve as a useful addition to clinical algorithms for determining the risk of and treatment stratification for hematoma expansion.

Project description:BackgroundNoncontrast Computed Tomography (NCCT) features are promising markers for acute hematoma expansion (HE) in patients with intracerebral hemorrhage (ICH). It remains unclear whether accurate identification of these markers is also reliable in raters with different levels of experience.MethodsPatients with acute spontaneous ICH admitted at four tertiary centers in Germany and Italy were retrospectively included from January 2017 to June 2020. In total, nine NCCT markers were rated by one radiology resident, one radiology fellow, and one neuroradiology fellow with different levels experience in ICH imaging. Interrater reliabilities of the resident and radiology fellow were evaluated by calculated Cohen's kappa (κ) statistics in reference to the neuroradiology fellow who was referred as the gold standard. Gold-standard ratings were evaluated by calculated interrater κ statistics. Global interrater reliabilities were evaluated by calculated Fleiss kappa statistics across all three readers. A comparison of receiver operating characteristics (ROCs) was used to evaluate differences in the diagnostic accuracy for predicting acute hematoma expansion (HE) among the raters.ResultsSubstantial-to-almost-perfect interrater concordance was found for the resident with interrater Cohen's kappa from 0.70 (95% CI 0.65-0.81) to 0.96 (95% CI 0.94-0.98). The interrater Cohen's kappa for the radiology fellow was moderate to almost perfect and ranged from 0.58 (95% CI 0.52-0.65) to 94 (95% CI 92-0.97). The intrarater gold-standard Cohen's kappa was almost perfect and ranged from 0.79 (95% CI 0.78-0.90) to 0.98 (95% CI 0.78-0.90). The global interrater Fleiss kappa ranged from 0.62 (95%CI 0.57-0.66) to 0.93 (95%CI 0.89-0.97). The diagnostic accuracy for the prediction of acute hematoma expansion (HE) was different for the island sign and fluid sign, with p-values < 0.05.ConclusionThe NCCT markers had a substantial-to-almost-perfect interrater agreement among raters with different levels of experience. Differences in the diagnostic accuracy for the prediction of acute HE were found in two out of nine NCCT markers. The study highlights the promising utility of NCCT markers for acute HE prediction.

Dataset Information

Uncertainty-aware deep-learning model for prediction of supratentorial hematoma expansion from admission non-contrast head computed tomography scan.

Publications

Uncertainty-aware deep-learning model for prediction of supratentorial hematoma expansion from admission non-contrast head computed tomography scan.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets